lkml.org 
[lkml]   [2000]   [Feb]   [15]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: lowlatency-2.2.14-B1 + 2.2.14aa7 fixes crash, but...

On Tue, 15 Feb 2000, William Montgomery wrote:

> all FAILED hunks except in entry.S for the following code snippit:
> -----part of entry.S.rej-------
> testl $(VM_MASK),EFLAGS(%esp)
> movl %esp,%eax
> jne v86_signal_return
> + #ifndef CONFIG_KERNEL_DEBUGGING
> xorl %edx,%edx
> + #else
> + pushl $0
> + pushl %eax
> + #endif
> call SYMBOL_NAME(do_signal)
> + #ifdef CONFIG_KERNEL_DEBUGGING
> + addl $8,%esp
> + #endif

> I think I can fix the FAILED hunk but I would feel more confident if I
> knew what this code was doing. Can you give a brief explanation of
> the ikd additions in this area?

thats a workaround i added ages ago - we switch off the FASTCALL macro
(regparm calling-convention extension of GCC which passes up to 3
parameters through %eax, %edx, %ecx) - in the IKD case the parameters go
through the stack, in the non-IKD case edx (second parameter) is zeroed,
eax is passed in as-is.

IKD triggered problems (likely bugs in previous GCC versions) if regparm
was used together with -pg (GCC's function-prologue profiling extension
calling _mcount). mcount() is the basis of the majority of IKD
'anti-lockup' and tracing features.

Ingo


-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:56    [from the cache]
©2003-2014 Jasper Spaans. hosted at Digital Ocean